早期胃癌におけるm・sm鑑別診断の現状

Accuracy of Preoperative Diagnosis of Early Gastric Cancer on Submucosal Invasion: A Study for Indications of Endoscopic Treatment

要旨　1986年から1990年までに当院で切除された長径2cm以下の単発性早期胃癌276例のうち，術前カンファレンスにおける最終診断の内容，および術後の病理記録が残されていた250例を対象として術前深達度診断の精度を検討した．全対象例の正診率は73％（182/250）であったが，術前診断別にみると，mと診断した189例の正診率は81％であったのに対し，smと診断した54例の正診率は54％と有意に不良であった（p＜0.01）．また，粘膜下層を均等に3分割し，sm浸潤の先進部の深さをsm₁からsm₃までの3段階に亜分類すると，mと誤診した36例のsm浸潤の程度は，sm₁ 12例，sm₂ 16例，sm₃ 8例であり，浅いものよりむしろ中間程度から深部に及ぶもののほうが多数を占めた，肉眼型別に正診率を比較すると，Ⅲあるいはその複合型（陥凹型）で極めて不良であったが，隆起型（ⅠおよびⅡa），表面陥凹型（Ⅱc，Ⅱbおよびその複合型），表面複合型（Ⅱa＋Ⅱc，Ⅱc＋Ⅱa）では，術前にmと診断した場合の正診率は80～90％と比較的良好であった．また，当院にて内視鏡的切除を行った45例の術前診断をみると，mとした41例中3例（7％）に予想外のsm浸潤を認めた．術前mとした誤診例25例について見直し診断を試みたが，sm浸潤を疑えたものは7例（28％）にすぎず，また見直し診断成績とsm浸潤量には特定の関連性を指摘しえなかった．以上により，2cm以下の早期胃癌に対する深達度診断を完壁に行うことはまず不可能と結論せざるを得なかった．しかし，術前にmと診断した隆起型，表面陥凹型，表面複合型の3者については80～90％の正診率が得られており，更に，わずかではあるが見直し診断で正診率を上乗せしえたことを考え合わせれば，現状の診断成績は内視鏡治療の適応を考えるうえでは十分に信頼に足りうると評価できる．

　To determine the appropriateness of endoscopic treatment of early gastric cancer (EGC), vertical invasivity should be evaluated before treatment, to exclude the possibility of lymph node metastasis. Preoperative diagnostic accuracy was evaluated on 250 of 276 consecutive cases of small (less than 2 cm) solitary EGC which had been resected surgically or endoscopically removed at NCCH during the period between 1986 and 1990.

　Of the 250 lesions, the preoperative diagnosis of 189 lesions was mucosal cancer and that of 54 lesions was submucosal or deeper invasive cancers. Accuracy rates of preoperative diagnoses were 73% for all cases, 81% for "mucosal cancer", and 54% for "submucosal cancer". There was significant difference (p＜0.01) between the latter two average rates. To investigate diagnostic accuracy of submucosal invasivity, tumors were divided into three groups depending on the extent of the submucosal invasion as follows; sm₁ as invasion within a third of submucosal layer, sm₂ as invasion involving up to two thirds, and sm₃ as invasion of greater than two thirds of submucosal layer. There was no statistical significance among these subgroups, however.

　In elevated, superficial depressed and superficial complex types, the accuracy rate of preoperative diagnosis of mucosal cancer was good (range 80-90%), whereas it was poor (60%) in the depressed type. Twenty five cases of submucosal cancer which were misdiagnosed as mucosal cancer by endoscopic examination were reviewed retrospectively, and seven cases (28%) could have been diagnosed as submucosal cancer; there was no relation between degree of submucosal invasivity and misdiagnosis.

　In conclusion, it is still impossible to estimate the invasivity of EGC perfectly, even in small lesions. Histological subclassification of submucosal invasivity is not helpful in making the correct diagnosis. The high accuracy rate of preoperative diagnosis in elevated, superficial depressed and superficial complex types of mucosal cancer may indicate that these lesions can be treated endoscopically.